we have 8 voices of polyphony per track 😊
Press unisono ,Select nr of voices , dial up detune spread dial , done
Also mixed polyphony/unisono like in KORG synth apps.
its gonna blow the house down ^^
It will be here soon, but the solution is different. An Unisono module - takes Pitch, Gate and Velocity CV. and spread them to available pool of voices.
Knobs: Num voices, Pitch spread, Delay spread, Velocity Spread + Randomize
Voices are reallocated dynamically.. e.g. if you set 4 unisono voices in module it will take only 2 voices from its input
@giku Excellent! 😊
@giku thats really dope, covers my needs!
Someday it would be cool to be able to “spread” any parameter in a poly/unison patch. Not necessary for my current plans but would make for some wild experiments. Maybe there is some way to do it currently?
Edit: With poly you can use keyboard as a modulator, not sure about unison tho. Maybe more effort than it’s worth.
Edit 2: I think I need to explore the voice # module :)
@giku I'd be interested in helping beta test this if there is need
update with Unisono is in the store now
@giku I've run into some issues on 1.24 that might be unison related. After using the unison module for a bit, and setting some modulations, I deleted it. Now when I play the patch monophonically there is variation in sound depending on whether Midi-to-cv is set to 1 voice or more.
The first part of the you can see how in the patch, its playing multiple voices even though there aren't multiple midi notes. The second part of the video I pulled oscillator from that patch into a separate track and trigger it. With more voices, it will only trigger 1/voices.
Attached is the patch, hope this help!
p.s. I had to zip it because the forum doesn't allow .drmodule files.
@quartzite A couple of us patch builders have stumbled over the not really intuitive polyphony handling in Drambo. If we have a MIDI 2 CV, a rack and a Layers module, we would expect the patch to "just work" when adding an envelope at the end. But sometimes it just doesn't, and the problem is that there's no hint at why tha is or what I should to do to make it work. Just because of this reason, I'm sometimes trying to avoid using racks and layers when working with polyphony.
You may spread any parameter in poly/unison. Just modulate it with Spread / Voice number output from Unison module. There is an extra Voice Number module.
I need more info :)
The only possible reason is using a monophonic module (e.g. reverb) before envelopes.
Thanks, checking this
Instrument rack outputs mono 1 voice signal (its by design). Yes its my fault, it isn't marked as a mono module.
So we better don't use Instrument racks in the Layers module? I've tried getting around this by routing pitch, gate and velocity signals from outside the racks, even from a new polyphonic MIDI to CV module post all the racks, none of them worked.
I thought I was only stumbling over voice stealing
I will add a protection against it.
uh than it makes no sense to use instrument racks as building blocks
uh uh uh
what a monday :)
I've had issues with this too. Maybe a Group Rack that just works keeping modules together and has configurable inputs/outputs.
Thanks for all this info and discussion! To analyze what's going on in my video/patch (I think?):
If I want to design a more complicated synth polyphonic synth, what is the best way?
Should I be putting the midi-to-cv inside my racks? Should I not be using racks at all? What's the best way to pack together discrete "units" in a patch (like my custom oscillators with ~8 modules)? What other tips?
I'm guessing the reason racks mix down output and strip the voice number info into Mono output... Is because the nature of modular construction allows for multiple branchings of signal to be processed in very different ways. The inclusion of CV/Gate sequencers, Math, Delay, etc... mean that there can be a multitude of various "poly signal mappings" that don't all align.
A way around this might be to create new module that is fundamentally just a segment of rack that allows anything within it to be interconnected in the usual way. It would essentially be the same thing as what the Section Modules do currently..... Simply delineate the boundaries of a section of rack between markers.
A Module based on a "Rack Segment", might hypothetically provide features for being "Compacted" or "Saved". But it would have a container type that has no connectivity associated with the container itself.
It would be up to the user to "remember" how to re-make connections any time they re-use any "Rack Segment" they had saved.
@quartzite What I've deciphered from @giku's response is that the instrument rack will "downmix" a polyphonic signal to mono so whatever you put after it will not be able to extract a polyphonic signal from its output, even if you add polyphonic gate and velocity signals by adding a new MIDI to CV module.
Your ADSR needs all signals in their original polyphonic form though, so from my understanding, you have two options:
In addition to the good info @rs2000 provided. Below is a photo example of a type of design method I use to preserve polyphony in synth designs:
The above Instrument Rack contains: ( Midi to CV - ENV AD - Processor Rack - Processor Rack - Mixer ).
Each "example" Processor Rack contains an ( Oscillator and an AMP ).
The key aspect of this type of design is putting a ENV AD in front of the Processor Racks... then sending the ENV CV into both Processor Racks. The ENV AD acts as a "remote controller" for the simple AMP's that are inside the Processor Racks. This type of configuration avoids putting an AMP ENV at the back, and outside the racks, where it can't function polyphonically. (Note there is a difference between an ENV AD and an AMP AD).
The technique of using "remote controllers" can be used in front of Processor Racks and Layer Racks.
I can build several different synths on each layer of a Layer Mixer, put a Slider Module in front of the Layer Mixer, and use that one external slider to control the filter cutoffs for every synth in the Layer Mixer.
This method is useful, but it has it's limits if attempting to build very complex synths.
I'm still looking for methods of constructing complex synths in Drambo "that work best for what I'd like to do". The latest addition of the modulatable Mixer Module, and the "CV router workaround technique" that rs2000 came up with, should be helpful.
Thanks y'all! I'll definitely refer to this next time I'm building up a patch :)